CN107018211A - A kind of monitoring method of whole machine cabinet server node information - Google Patents

A kind of monitoring method of whole machine cabinet server node information Download PDF

Info

Publication number
CN107018211A
CN107018211A CN201710342730.3A CN201710342730A CN107018211A CN 107018211 A CN107018211 A CN 107018211A CN 201710342730 A CN201710342730 A CN 201710342730A CN 107018211 A CN107018211 A CN 107018211A
Authority
CN
China
Prior art keywords
bmc
node
rmc
network
whole machine
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201710342730.3A
Other languages
Chinese (zh)
Other versions
CN107018211B (en
Inventor
苏孝
班华堂
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhengzhou Yunhai Information Technology Co Ltd
Original Assignee
Zhengzhou Yunhai Information Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Zhengzhou Yunhai Information Technology Co Ltd filed Critical Zhengzhou Yunhai Information Technology Co Ltd
Priority to CN201710342730.3A priority Critical patent/CN107018211B/en
Publication of CN107018211A publication Critical patent/CN107018211A/en
Application granted granted Critical
Publication of CN107018211B publication Critical patent/CN107018211B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L61/00Network arrangements, protocols or services for addressing or naming
    • H04L61/50Address allocation
    • H04L61/5046Resolving address allocation conflicts; Testing of addresses
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00Arrangements for monitoring or testing data switching networks
    • H04L43/08Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
    • H04L43/0805Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters by checking availability
    • H04L43/0817Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters by checking availability by checking functioning
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L61/00Network arrangements, protocols or services for addressing or naming
    • H04L61/50Address allocation
    • H04L61/5053Lease time; Renewal aspects
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L69/00Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass
    • H04L69/16Implementation or adaptation of Internet protocol [IP], of transmission control protocol [TCP] or of user datagram protocol [UDP]
    • H04L69/161Implementation details of TCP/IP or UDP/IP stack architecture; Specification of modified or new header fields
    • H04L69/162Implementation details of TCP/IP or UDP/IP stack architecture; Specification of modified or new header fields involving adaptations of sockets based mechanisms

Abstract

The present invention relates to server info monitoring field, and in particular to a kind of monitoring method of whole machine cabinet server node information.Plate BMC three-level communication links in RMC nodes are reduced to RMC BMC two-stage communication links by the monitoring method, greatly accelerate RMC and node BMC communication process, RMC quickly can get nodal information especially real-time change information from BMC, reduce system response time, just rotation speed of the fan can be adjusted according to current heat dissipating state quickly in node load suddenly change, radiating effect is improved, while can also substantially reduce fan wall and node power consumption.

Description

A kind of monitoring method of whole machine cabinet server node information
Technical field
The present invention relates to server info monitoring field, and in particular to a kind of monitoring side of whole machine cabinet server node information Method.Plate-BMC three-level communication links in RMC- nodes are reduced to RMC-BMC two-stage communication links by the monitoring method, are greatly speeded up RMC and node BMC communication process, RMC quickly can get nodal information especially real-time change information from BMC, reduces System response time, just rotation speed of the fan can be adjusted according to current heat dissipating state quickly in node load suddenly change, improve and dissipate Thermal effect, while can also substantially reduce fan wall and node power consumption.
Background technology
Network-node server (Network Node Server, NNS) is SNA network node (Network Node) A kind of server unit.For node server is for server cluster, it is mainly used in WEB, FTP, VPN etc. clothes In business.In whole machine cabinet server system, whole machine cabinet administrative unit RMC (Rack Management Controller, rack pipe Manage unit) (Baseboard Management Controller, substrate management is controlled with node server administrative unit BMC Device) between communication process be usually so:In node plate by I2C interfaces from node BMC obtain this node layer in plate connect The several node server information connect, RMC by I2C interfaces from each node plate indirect gain to node server information, from And be used to be monitored whole machine cabinet server node information management and rotation speed of the fan control etc..
In node between plate and node BMC, to be communicated in RMC and node between plate communicated by I2C interfaces, I2C communications are unstable, and speed is slower, and antijamming capability is poor, and whole machine cabinet fan speed-regulating policy mandates RMC in itself can be quick The temperature information of node is collected, in node server pressure jump, cpu temperature rapid increase, because traffic rate is slow, newly Temperature information the long period is needed from BMC to RMC, it is impossible to meet the ageing high request of complete machine cabinet heat radiation well, cause Radiating effect is undesirable.On the other hand, communicated using I2C, whole machine cabinet needs to use plate-node BMC triodes in RMC- nodes Manage module communication, three-level management module also increases the time delay of information transfer, be unfavorable for whole machine cabinet nodal information quickly transmission and Speed adjustment strategy quick response.
In view of the above-mentioned problems, the present application a kind of monitoring method of whole machine cabinet server node information, whole machine cabinet pipe The communication mode that reason unit R MC and BMC will be combined using I2C and network, by plate-BMC three-level communication links in RMC- nodes RMC-BMC two-stage communication links are reduced to, not only overcome that I2C communications speed itself is slower, unstable, poor anti jamming capability Shortcoming, and greatly accelerate RMC and node BMC communication process, RMC quickly can get nodal information from BMC, especially Real-time change information, it is of great advantage to improving system response time.For complete machine cabinet heat radiation speed governing, because RMC can be fast from node Speed gets the information such as CPU, internal memory, intake air temperature, just can be quickly according to current radiating shape in node load suddenly change Condition adjusts rotation speed of the fan, radiating effect is improved, while can also substantially reduce fan wall and node power consumption.
The content of the invention
Invention described herein proposes a kind of monitoring method of whole machine cabinet server node information, and whole machine cabinet management is single The communication mode that first RMC and BMC will be combined using I2C and network, plate-BMC three-levels communication link in RMC- nodes is simplified For RMC-BMC two-stage communication links, concrete technical scheme is as follows:
A kind of monitoring method of whole machine cabinet server node information, the monitoring method specifically includes following steps:Start section Plate and RMC software in point;RMC obtains all node BMC of whole machine cabinet dedicated port and shared mouth network ip address;RMC judges BMC Can dedicated port IP connect:If can connect, RMC carries out socket communications by BMC dedicated ports network and BMC, obtains server Node related information;If BMC dedicated ports network can not be connected, RMC continues to judge that can BMC share mouth network and connect:If energy Connection, RMC shares mouth network by BMC and BMC carries out socket communications;If can not connect, RMC continues through I2C from each layer Plate obtains node server information in node.
The monitoring method of whole machine cabinet server node information as described above, is further characterized in that, RMC obtains whole machine cabinet institute The step of having node BMC dedicated port and shared mouth network ip address specifically includes:Plate obtains this by I2C interfaces in node Plate connects the IP address of several node BMC dedicated ports networks and shared mouth network in node layer, and RMC is from every node layer between plate Obtain the dedicated port and shared mouth network ip address for getting all node BMC of whole machine cabinet.
The monitoring method of whole machine cabinet server node information as described above, is further characterized in that, RMC and node BMC In Socket communication process, if BMC dedicated port network interruptions, RMC continues to judge shared mouth network-in-dialing state and according to net Network connected state decides whether to proceed socket communications by shared mouth.
The monitoring method of whole machine cabinet server node information as described above, is further characterized in that, RMC and node BMC In Socket communication process, if BMC dedicated ports network is interrupted and can not connected with shared mouth network, RMC be continuing with I2C from Plate obtains node server information in each node layer.
The monitoring method of whole machine cabinet server node information as described above, is further characterized in that, if RMC and node BMC In Socket communication process, BMC network ip address is changed, then network service is interrupted, RMC plates from node again through I2C New IP address is got, network service is directly carried out by new IP address and node BMC.
Brief description of the drawings
Fig. 1, monitoring method of the present invention workflow diagram
Embodiment
In invention described herein, whole machine cabinet administrative unit RMC and node server administrative unit BMC communications are followed The rule of communication of Network Priority, i.e., when RMC and BMC dedicated ports network or shared mouth network-in-dialing, preferentially by network service, When RMC with BMC dedicated ports network and shared mouth network can not be connected, then by I2C from node plate indirect gain node Server info.Detailed process is as follows:
(1) software starts after electricity on plate in RMC, node, and plate obtains institute's connecting node BMC dedicated ports by I2C in node Network and shared mouth network ip address, RMC plates from node get all node BMC of whole machine cabinet dedicated port and shared mouth net Network IP address.
(2) RMC judges that can BMC dedicated ports connect with shared mouth IP:It is special by BMC if dedicated port network can be connected Mouth network service is used, if BMC dedicated ports network can not be connected, RMC judges that can BMC share mouth network and connect, if can connect, RMC shares mouth network by BMC and BMC carries out socket communications, if can not connect, RMC continues through I2C plates from node and obtained Take node server information.
(3) during RMC and node BMC network services, if in the BMC dedicated ports network or shared mouth network that use Disconnected, then RMC continues to judge and attempt by being communicated currently without the shared mouth or dedicated port that use.
(4) during RMC and node BMC network services, if BMC dedicated ports network and shared mouth network are interrupted, RMC It is continuing with I2C interfaces plate from node and obtains node server information, untill BMC network-in-dialing.
Whole machine cabinet administrative unit RMC and node server administrative unit BMC communication modes described in the invention, not only gram Take that I2C communications speed itself is slower, unstable, poor anti jamming capability shortcoming, and by plate-BMC three-levels in RMC- nodes Communication link is reduced to RMC-BMC two-stage communication links, greatly accelerates RMC and node BMC communication process, RMC can quickly from BMC gets nodal information, especially real-time change information, of great advantage to improving system response time.Dissipated for whole machine cabinet Hot speed governing, because RMC can be from node quick obtaining to CPU, the information such as internal memory, intake air temperature, in node load suddenly change When just can quickly according to current heat dissipating state adjust rotation speed of the fan, improve radiating effect, while can also substantially reduce fan wall and Node power consumption.

Claims (5)

1. a kind of monitoring method of whole machine cabinet server node information, the monitoring method specifically includes following steps:
Plate and RMC software in starter node;
RMC obtains all node BMC of whole machine cabinet dedicated port and shared mouth network ip address;
RMC judges that can BMC dedicated ports IP connect:
If can connect, RMC carries out socket communications by BMC dedicated ports network and BMC, obtains server node relevant information;
If BMC dedicated ports network can not be connected, RMC continues to judge that can BMC share mouth network and connect:
If can connect, RMC shares mouth network by BMC and BMC carries out socket communications;
If can not connect, RMC continues through I2C plates from each node layer and obtains node server information.
2. the monitoring method of the whole machine cabinet server node information described in claim 1, is further characterized in that, RMC obtains complete machine The step of all node BMC of cabinet dedicated port and shared mouth network ip address, specifically includes:
Plate obtains plate in this node layer by I2C interfaces and connects several node BMC dedicated ports networks and shared mouth net in node The IP address of network, the RMC dedicated ports and shared mouth network IP of plate indirect gain to all node BMC of whole machine cabinet from every node layer Address.
3. the monitoring method of whole machine cabinet server node information as claimed in claim 2, is further characterized in that, RMC and node In BMC Socket communication process, if BMC dedicated port network interruptions, RMC continues to judge shared mouth network-in-dialing state and root Communicated according to whether network-in-dialing Determines proceed socket by shared mouth.
4. the monitoring method of whole machine cabinet server node information as claimed in claim 3, is further characterized in that, RMC and node In BMCSocket communication process, if BMC dedicated ports network is interrupted and can not connected with shared mouth network, RMC is continuing with I2C Plate obtains node server information from each node layer.
5. the monitoring method of whole machine cabinet server node information as claimed in claim 4, is further characterized in that, if RMC and section In point BMC Socket communication process, BMC network ip address is changed, then network service is interrupted, and RMC is again through I2C from section Plate gets new IP address in point, and network service is directly carried out by new IP address and node BMC.
CN201710342730.3A 2017-05-15 2017-05-15 Method for monitoring node information of server of whole cabinet Active CN107018211B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710342730.3A CN107018211B (en) 2017-05-15 2017-05-15 Method for monitoring node information of server of whole cabinet

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710342730.3A CN107018211B (en) 2017-05-15 2017-05-15 Method for monitoring node information of server of whole cabinet

Publications (2)

Publication Number Publication Date
CN107018211A true CN107018211A (en) 2017-08-04
CN107018211B CN107018211B (en) 2020-03-24

Family

ID=59449997

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710342730.3A Active CN107018211B (en) 2017-05-15 2017-05-15 Method for monitoring node information of server of whole cabinet

Country Status (1)

Country Link
CN (1) CN107018211B (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107302465A (en) * 2017-08-18 2017-10-27 郑州云海信息技术有限公司 A kind of PCIe Switch servers complete machine management method
CN108540551A (en) * 2018-04-04 2018-09-14 郑州云海信息技术有限公司 A kind of acquisition methods of server node information and obtain system
CN108768693A (en) * 2018-04-19 2018-11-06 华勤通讯技术有限公司 The recognition methods of shared port, storage medium and server in Multi net voting port
CN113645067A (en) * 2021-08-03 2021-11-12 加弘科技咨询(上海)有限公司 Network management system and method of multi-node substrate management controller

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105373462A (en) * 2015-11-12 2016-03-02 浪潮(北京)电子信息产业有限公司 Whole cabinet server management method and system
CN105577430A (en) * 2015-12-15 2016-05-11 山东海量信息技术研究院 Node management method of high-end fault-tolerant server
CN105681359A (en) * 2016-04-01 2016-06-15 浪潮电子信息产业股份有限公司 Rack monitoring device and method
US20160239370A1 (en) * 2015-02-12 2016-08-18 Aic Inc. Rack having automatic recovery function and automatic recovery method for the same
CN106598183A (en) * 2016-12-26 2017-04-26 郑州云海信息技术有限公司 Two-stage fan regulation and control system and method applicable to multi-node server

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160239370A1 (en) * 2015-02-12 2016-08-18 Aic Inc. Rack having automatic recovery function and automatic recovery method for the same
CN105373462A (en) * 2015-11-12 2016-03-02 浪潮(北京)电子信息产业有限公司 Whole cabinet server management method and system
CN105577430A (en) * 2015-12-15 2016-05-11 山东海量信息技术研究院 Node management method of high-end fault-tolerant server
CN105681359A (en) * 2016-04-01 2016-06-15 浪潮电子信息产业股份有限公司 Rack monitoring device and method
CN106598183A (en) * 2016-12-26 2017-04-26 郑州云海信息技术有限公司 Two-stage fan regulation and control system and method applicable to multi-node server

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107302465A (en) * 2017-08-18 2017-10-27 郑州云海信息技术有限公司 A kind of PCIe Switch servers complete machine management method
CN108540551A (en) * 2018-04-04 2018-09-14 郑州云海信息技术有限公司 A kind of acquisition methods of server node information and obtain system
CN108768693A (en) * 2018-04-19 2018-11-06 华勤通讯技术有限公司 The recognition methods of shared port, storage medium and server in Multi net voting port
CN108768693B (en) * 2018-04-19 2021-07-23 华勤技术股份有限公司 Method for identifying shared port in multiple network ports, storage medium and server
CN113645067A (en) * 2021-08-03 2021-11-12 加弘科技咨询(上海)有限公司 Network management system and method of multi-node substrate management controller

Also Published As

Publication number Publication date
CN107018211B (en) 2020-03-24

Similar Documents

Publication Publication Date Title
CN107018211A (en) A kind of monitoring method of whole machine cabinet server node information
US10003540B2 (en) Flow forwarding method, device, and system
EP2922254B1 (en) Method for establishing control channel, forwarding point and controller
CN104618194A (en) Software definition network message monitoring method, SDN controller and switching devices
EP3567812B1 (en) Method, device, and system for adjusting a forwarding path
US9743367B2 (en) Link layer discovery protocol (LLDP) on multiple nodes of a distributed fabric
US11595390B2 (en) Self-organizing trusted networks
US9960998B2 (en) Forwarding packet in stacking system
WO2022001669A1 (en) Method for establishing vxlan tunnel, and related device
CN111464454B (en) Load sharing method and system for virtual BRAS (broadband remote Access Server) equipment in data center
TWI639325B (en) Automatically configured switch,method of automatically configuring a switch, and software defined network system with auto-deployment switches and auto-deploying method thereof
WO2015180154A1 (en) Network control method and apparatus
US20110176551A1 (en) Information Handling System Data Center Bridging Features with Defined Application Environments
WO2021047448A1 (en) Session management method and device
CN107204907B (en) Cloud data center interconnection method and device
US11140091B2 (en) Openflow protocol-based resource control method and system, and apparatus
CN105191339A (en) Software redundancy in a non-homogenous virtual chassis
US11223597B2 (en) Network and network management method
WO2015139359A1 (en) Wireless network maintenance method, device and system
JP2011041251A (en) Architecture using inexpensively managed radio switching point for distribution of large scale wireless lan
WO2017015839A1 (en) Sdn-based arp realization method and apparatus
TW201340642A (en) Remote server managing system and remote server managing method
WO2013189069A1 (en) Load sharing method and device, and single board
EP4214911A1 (en) Systems and methods for zero-touch provisioning of a switch in intermediate distribution frames and main distribution frames
WO2015188331A1 (en) 转发控制方法、驱动器及sdn网络 forwarding control method, driver and sdn network

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant